[v2] seqlock: introduce scoped_seqlock_read() and scoped_seqlock_read_irqsave()

[PATCH v2 0/4] seqlock: introduce scoped_seqlock_read() and scoped_seqlock_read_irqsave()

Posted by Oleg Nesterov 4 months ago

Only 1/4 was changed according to comments from Linus and Waiman,
but let me resend 2-4 as well.

Oleg.

 fs/proc/array.c         |  9 ++-----
 fs/proc/base.c          | 10 ++------
 include/linux/seqlock.h | 64 +++++++++++++++++++++++++++++++++++++++++++++++++
 kernel/sched/cputime.c  | 15 +++---------
 4 files changed, 71 insertions(+), 27 deletions(-)

Re: [PATCH v2 0/4] seqlock: introduce scoped_seqlock_read() and scoped_seqlock_read_irqsave()

Posted by Peter Zijlstra 4 months ago

On Wed, Oct 08, 2025 at 02:30:14PM +0200, Oleg Nesterov wrote:
> Only 1/4 was changed according to comments from Linus and Waiman,
> but let me resend 2-4 as well.

Shall I put this in tip/locking/core once -rc1 happens or so?

Re: [PATCH v2 0/4] seqlock: introduce scoped_seqlock_read() and scoped_seqlock_read_irqsave()

Posted by Oleg Nesterov 4 months ago

On 10/08, Peter Zijlstra wrote:
>
> On Wed, Oct 08, 2025 at 02:30:14PM +0200, Oleg Nesterov wrote:
> > Only 1/4 was changed according to comments from Linus and Waiman,
> > but let me resend 2-4 as well.
>
> Shall I put this in tip/locking/core once -rc1 happens or so?

Would be great ;) If nobody objects.

Can you also take the trivial

	[PATCH 1/1] documentation: seqlock: fix the wrong documentation of read_seqbegin_or_lock/need_seqretry
	https://lore.kernel.org/all/20250928162029.GA3121@redhat.com/

?

OTOH, if this series is merged, this doc fix should be updated to
mention scoped_seqlock_read()...

Oleg.

Re: [PATCH v2 0/4] seqlock: introduce scoped_seqlock_read() and scoped_seqlock_read_irqsave()

Posted by Peter Zijlstra 4 months ago

On Wed, Oct 08, 2025 at 03:13:02PM +0200, Oleg Nesterov wrote:
> On 10/08, Peter Zijlstra wrote:
> >
> > On Wed, Oct 08, 2025 at 02:30:14PM +0200, Oleg Nesterov wrote:
> > > Only 1/4 was changed according to comments from Linus and Waiman,
> > > but let me resend 2-4 as well.
> >
> > Shall I put this in tip/locking/core once -rc1 happens or so?
> 
> Would be great ;) If nobody objects.
> 
> Can you also take the trivial
> 
> 	[PATCH 1/1] documentation: seqlock: fix the wrong documentation of read_seqbegin_or_lock/need_seqretry
> 	https://lore.kernel.org/all/20250928162029.GA3121@redhat.com/
> 
> ?

I've picked that up too and stuck it in front.

> OTOH, if this series is merged, this doc fix should be updated to
> mention scoped_seqlock_read()...

I see a patch [5/4] in your future ;-)